[jira] [Updated] (HBASE-4899) Region would be assigned twice easily with continually killing server and moving region in testing environment
[ https://issues.apache.org/jira/browse/HBASE-4899?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] ramkrishna.s.vasudevan updated HBASE-4899: -- Affects Version/s: (was: 0.92.1) 0.92.0 Fix Version/s: (was: 0.90.0) 0.92.0 > Region would be assigned twice easily with continually killing server and > moving region in testing environment > --- > > Key: HBASE-4899 > URL: https://issues.apache.org/jira/browse/HBASE-4899 > Project: HBase > Issue Type: Bug >Affects Versions: 0.92.0 >Reporter: chunhui shen >Assignee: chunhui shen >Priority: Critical > Fix For: 0.92.0 > > Attachments: hbase-4899.patch, hbase-4899v2.patch, hbase-4899v3.patch > > > Before assigning region in ServerShutdownHandler#process, it will check > whether region is in RIT, > however, this checking doesn't work as the excepted in the following case: > 1.move region A from server B to server C > 2.kill server B > 3.start server B immediately > Let's see what happen in the code for the above case > {code} > for step1: > 1.1 server B close the region A, > 1.2 master setOffline for region > A,(AssignmentManager#setOffline:this.regions.remove(regionInfo)) > 1.3 server C start to open region A.(Not completed) > for step3: > master ServerShutdownHandler#process() for server B > { > .. > splitlog() > ... > List regionsInTransition = > this.services.getAssignmentManager() > .processServerShutdown(this.serverName); > ... > Skip regions that were in transition unless CLOSING or PENDING_CLOSE > ... > assign region > } > {code} > In fact, when running > ServerShutdownHandler#process()#this.services.getAssignmentManager().processServerShutdown(this.serverName), > region A is in RIT (step1.3 not completed), but the return List > regionsInTransition doesn't contain it, because region A has removed from > AssignmentManager.regions by AssignmentManager#setOffline in step 1.2 > Therefore, region A will be assigned twice. > Actually, one server killed and started twice will also easily cause region > assigned twice. > Exclude the above reason, another probability : > when execute ServerShutdownHandler#process()#MetaReader.getServerUserRegions > ,region is included which is in RIT now. > But after completing MetaReader.getServerUserRegions, the region has been > opened in other server and is not in RIT now. > In our testing environment where balancing,moving and killing are executed > periodly, assigning region twice often happens, and it is hateful because it > will affect other test cases. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4899) Region would be assigned twice easily with continually killing server and moving region in testing environment
[ https://issues.apache.org/jira/browse/HBASE-4899?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] stack updated HBASE-4899: - Fix Version/s: (was: 0.92.1) 0.90.0 This will be in 0.92 afterall; cutting a new RC. > Region would be assigned twice easily with continually killing server and > moving region in testing environment > --- > > Key: HBASE-4899 > URL: https://issues.apache.org/jira/browse/HBASE-4899 > Project: HBase > Issue Type: Bug >Affects Versions: 0.92.1 >Reporter: chunhui shen >Assignee: chunhui shen >Priority: Critical > Fix For: 0.90.0 > > Attachments: hbase-4899.patch, hbase-4899v2.patch, hbase-4899v3.patch > > > Before assigning region in ServerShutdownHandler#process, it will check > whether region is in RIT, > however, this checking doesn't work as the excepted in the following case: > 1.move region A from server B to server C > 2.kill server B > 3.start server B immediately > Let's see what happen in the code for the above case > {code} > for step1: > 1.1 server B close the region A, > 1.2 master setOffline for region > A,(AssignmentManager#setOffline:this.regions.remove(regionInfo)) > 1.3 server C start to open region A.(Not completed) > for step3: > master ServerShutdownHandler#process() for server B > { > .. > splitlog() > ... > List regionsInTransition = > this.services.getAssignmentManager() > .processServerShutdown(this.serverName); > ... > Skip regions that were in transition unless CLOSING or PENDING_CLOSE > ... > assign region > } > {code} > In fact, when running > ServerShutdownHandler#process()#this.services.getAssignmentManager().processServerShutdown(this.serverName), > region A is in RIT (step1.3 not completed), but the return List > regionsInTransition doesn't contain it, because region A has removed from > AssignmentManager.regions by AssignmentManager#setOffline in step 1.2 > Therefore, region A will be assigned twice. > Actually, one server killed and started twice will also easily cause region > assigned twice. > Exclude the above reason, another probability : > when execute ServerShutdownHandler#process()#MetaReader.getServerUserRegions > ,region is included which is in RIT now. > But after completing MetaReader.getServerUserRegions, the region has been > opened in other server and is not in RIT now. > In our testing environment where balancing,moving and killing are executed > periodly, assigning region twice often happens, and it is hateful because it > will affect other test cases. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4899) Region would be assigned twice easily with continually killing server and moving region in testing environment
[ https://issues.apache.org/jira/browse/HBASE-4899?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] stack updated HBASE-4899: - Resolution: Fixed Fix Version/s: 0.92.1 Hadoop Flags: Reviewed Status: Resolved (was: Patch Available) Applied to trunk and 0.92 branch. I ran all tests and first time a TestReplication failed. Its failing on trunk and 0.92 at mo. Second time I ran it all tests passed. Thanks for the patch Chunhui. > Region would be assigned twice easily with continually killing server and > moving region in testing environment > --- > > Key: HBASE-4899 > URL: https://issues.apache.org/jira/browse/HBASE-4899 > Project: HBase > Issue Type: Bug >Affects Versions: 0.92.1 >Reporter: chunhui shen >Assignee: chunhui shen >Priority: Critical > Fix For: 0.92.1 > > Attachments: hbase-4899.patch, hbase-4899v2.patch, hbase-4899v3.patch > > > Before assigning region in ServerShutdownHandler#process, it will check > whether region is in RIT, > however, this checking doesn't work as the excepted in the following case: > 1.move region A from server B to server C > 2.kill server B > 3.start server B immediately > Let's see what happen in the code for the above case > {code} > for step1: > 1.1 server B close the region A, > 1.2 master setOffline for region > A,(AssignmentManager#setOffline:this.regions.remove(regionInfo)) > 1.3 server C start to open region A.(Not completed) > for step3: > master ServerShutdownHandler#process() for server B > { > .. > splitlog() > ... > List regionsInTransition = > this.services.getAssignmentManager() > .processServerShutdown(this.serverName); > ... > Skip regions that were in transition unless CLOSING or PENDING_CLOSE > ... > assign region > } > {code} > In fact, when running > ServerShutdownHandler#process()#this.services.getAssignmentManager().processServerShutdown(this.serverName), > region A is in RIT (step1.3 not completed), but the return List > regionsInTransition doesn't contain it, because region A has removed from > AssignmentManager.regions by AssignmentManager#setOffline in step 1.2 > Therefore, region A will be assigned twice. > Actually, one server killed and started twice will also easily cause region > assigned twice. > Exclude the above reason, another probability : > when execute ServerShutdownHandler#process()#MetaReader.getServerUserRegions > ,region is included which is in RIT now. > But after completing MetaReader.getServerUserRegions, the region has been > opened in other server and is not in RIT now. > In our testing environment where balancing,moving and killing are executed > periodly, assigning region twice often happens, and it is hateful because it > will affect other test cases. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4899) Region would be assigned twice easily with continually killing server and moving region in testing environment
[ https://issues.apache.org/jira/browse/HBASE-4899?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] chunhui shen updated HBASE-4899: Attachment: hbase-4899v3.patch > Region would be assigned twice easily with continually killing server and > moving region in testing environment > --- > > Key: HBASE-4899 > URL: https://issues.apache.org/jira/browse/HBASE-4899 > Project: HBase > Issue Type: Bug >Affects Versions: 0.92.1 >Reporter: chunhui shen >Assignee: chunhui shen >Priority: Critical > Attachments: hbase-4899.patch, hbase-4899v2.patch, hbase-4899v3.patch > > > Before assigning region in ServerShutdownHandler#process, it will check > whether region is in RIT, > however, this checking doesn't work as the excepted in the following case: > 1.move region A from server B to server C > 2.kill server B > 3.start server B immediately > Let's see what happen in the code for the above case > {code} > for step1: > 1.1 server B close the region A, > 1.2 master setOffline for region > A,(AssignmentManager#setOffline:this.regions.remove(regionInfo)) > 1.3 server C start to open region A.(Not completed) > for step3: > master ServerShutdownHandler#process() for server B > { > .. > splitlog() > ... > List regionsInTransition = > this.services.getAssignmentManager() > .processServerShutdown(this.serverName); > ... > Skip regions that were in transition unless CLOSING or PENDING_CLOSE > ... > assign region > } > {code} > In fact, when running > ServerShutdownHandler#process()#this.services.getAssignmentManager().processServerShutdown(this.serverName), > region A is in RIT (step1.3 not completed), but the return List > regionsInTransition doesn't contain it, because region A has removed from > AssignmentManager.regions by AssignmentManager#setOffline in step 1.2 > Therefore, region A will be assigned twice. > Actually, one server killed and started twice will also easily cause region > assigned twice. > Exclude the above reason, another probability : > when execute ServerShutdownHandler#process()#MetaReader.getServerUserRegions > ,region is included which is in RIT now. > But after completing MetaReader.getServerUserRegions, the region has been > opened in other server and is not in RIT now. > In our testing environment where balancing,moving and killing are executed > periodly, assigning region twice often happens, and it is hateful because it > will affect other test cases. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4899) Region would be assigned twice easily with continually killing server and moving region in testing environment
[ https://issues.apache.org/jira/browse/HBASE-4899?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] stack updated HBASE-4899: - Priority: Critical (was: Major) Affects Version/s: (was: 0.92.0) 0.92.1 Upping priority and marking against 0.92.1. Will pull into 0.92.0 if another RC. > Region would be assigned twice easily with continually killing server and > moving region in testing environment > --- > > Key: HBASE-4899 > URL: https://issues.apache.org/jira/browse/HBASE-4899 > Project: HBase > Issue Type: Bug >Affects Versions: 0.92.1 >Reporter: chunhui shen >Assignee: chunhui shen >Priority: Critical > Attachments: hbase-4899.patch, hbase-4899v2.patch > > > Before assigning region in ServerShutdownHandler#process, it will check > whether region is in RIT, > however, this checking doesn't work as the excepted in the following case: > 1.move region A from server B to server C > 2.kill server B > 3.start server B immediately > Let's see what happen in the code for the above case > {code} > for step1: > 1.1 server B close the region A, > 1.2 master setOffline for region > A,(AssignmentManager#setOffline:this.regions.remove(regionInfo)) > 1.3 server C start to open region A.(Not completed) > for step3: > master ServerShutdownHandler#process() for server B > { > .. > splitlog() > ... > List regionsInTransition = > this.services.getAssignmentManager() > .processServerShutdown(this.serverName); > ... > Skip regions that were in transition unless CLOSING or PENDING_CLOSE > ... > assign region > } > {code} > In fact, when running > ServerShutdownHandler#process()#this.services.getAssignmentManager().processServerShutdown(this.serverName), > region A is in RIT (step1.3 not completed), but the return List > regionsInTransition doesn't contain it, because region A has removed from > AssignmentManager.regions by AssignmentManager#setOffline in step 1.2 > Therefore, region A will be assigned twice. > Actually, one server killed and started twice will also easily cause region > assigned twice. > Exclude the above reason, another probability : > when execute ServerShutdownHandler#process()#MetaReader.getServerUserRegions > ,region is included which is in RIT now. > But after completing MetaReader.getServerUserRegions, the region has been > opened in other server and is not in RIT now. > In our testing environment where balancing,moving and killing are executed > periodly, assigning region twice often happens, and it is hateful because it > will affect other test cases. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4899) Region would be assigned twice easily with continually killing server and moving region in testing environment
[ https://issues.apache.org/jira/browse/HBASE-4899?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] chunhui shen updated HBASE-4899: Attachment: hbase-4899v2.patch > Region would be assigned twice easily with continually killing server and > moving region in testing environment > --- > > Key: HBASE-4899 > URL: https://issues.apache.org/jira/browse/HBASE-4899 > Project: HBase > Issue Type: Bug >Affects Versions: 0.92.0 >Reporter: chunhui shen >Assignee: chunhui shen > Attachments: hbase-4899.patch, hbase-4899v2.patch > > > Before assigning region in ServerShutdownHandler#process, it will check > whether region is in RIT, > however, this checking doesn't work as the excepted in the following case: > 1.move region A from server B to server C > 2.kill server B > 3.start server B immediately > Let's see what happen in the code for the above case > {code} > for step1: > 1.1 server B close the region A, > 1.2 master setOffline for region > A,(AssignmentManager#setOffline:this.regions.remove(regionInfo)) > 1.3 server C start to open region A.(Not completed) > for step3: > master ServerShutdownHandler#process() for server B > { > .. > splitlog() > ... > List regionsInTransition = > this.services.getAssignmentManager() > .processServerShutdown(this.serverName); > ... > Skip regions that were in transition unless CLOSING or PENDING_CLOSE > ... > assign region > } > {code} > In fact, when running > ServerShutdownHandler#process()#this.services.getAssignmentManager().processServerShutdown(this.serverName), > region A is in RIT (step1.3 not completed), but the return List > regionsInTransition doesn't contain it, because region A has removed from > AssignmentManager.regions by AssignmentManager#setOffline in step 1.2 > Therefore, region A will be assigned twice. > Actually, one server killed and started twice will also easily cause region > assigned twice. > Exclude the above reason, another probability : > when execute ServerShutdownHandler#process()#MetaReader.getServerUserRegions > ,region is included which is in RIT now. > But after completing MetaReader.getServerUserRegions, the region has been > opened in other server and is not in RIT now. > In our testing environment where balancing,moving and killing are executed > periodly, assigning region twice often happens, and it is hateful because it > will affect other test cases. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4899) Region would be assigned twice easily with continually killing server and moving region in testing environment
[ https://issues.apache.org/jira/browse/HBASE-4899?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ted Yu updated HBASE-4899: -- Status: Patch Available (was: Open) > Region would be assigned twice easily with continually killing server and > moving region in testing environment > --- > > Key: HBASE-4899 > URL: https://issues.apache.org/jira/browse/HBASE-4899 > Project: HBase > Issue Type: Bug >Affects Versions: 0.92.0 >Reporter: chunhui shen >Assignee: chunhui shen > Attachments: hbase-4899.patch > > > Before assigning region in ServerShutdownHandler#process, it will check > whether region is in RIT, > however, this checking doesn't work as the excepted in the following case: > 1.move region A from server B to server C > 2.kill server B > 3.start server B immediately > Let's see what happen in the code for the above case > {code} > for step1: > 1.1 server B close the region A, > 1.2 master setOffline for region > A,(AssignmentManager#setOffline:this.regions.remove(regionInfo)) > 1.3 server C start to open region A.(Not completed) > for step3: > master ServerShutdownHandler#process() for server B > { > .. > splitlog() > ... > List regionsInTransition = > this.services.getAssignmentManager() > .processServerShutdown(this.serverName); > ... > Skip regions that were in transition unless CLOSING or PENDING_CLOSE > ... > assign region > } > {code} > In fact, when running > ServerShutdownHandler#process()#this.services.getAssignmentManager().processServerShutdown(this.serverName), > region A is in RIT (step1.3 not completed), but the return List > regionsInTransition doesn't contain it, because region A has removed from > AssignmentManager.regions by AssignmentManager#setOffline in step 1.2 > Therefore, region A will be assigned twice. > Actually, one server killed and started twice will also easily cause region > assigned twice. > Exclude the above reason, another probability : > when execute ServerShutdownHandler#process()#MetaReader.getServerUserRegions > ,region is included which is in RIT now. > But after completing MetaReader.getServerUserRegions, the region has been > opened in other server and is not in RIT now. > In our testing environment where balancing,moving and killing are executed > periodly, assigning region twice often happens, and it is hateful because it > will affect other test cases. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4899) Region would be assigned twice easily with continually killing server and moving region in testing environment
[ https://issues.apache.org/jira/browse/HBASE-4899?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ted Yu updated HBASE-4899: -- Description: Before assigning region in ServerShutdownHandler#process, it will check whether region is in RIT, however, this checking doesn't work as the excepted in the following case: 1.move region A from server B to server C 2.kill server B 3.start server B immediately Let's see what happen in the code for the above case {code} for step1: 1.1 server B close the region A, 1.2 master setOffline for region A,(AssignmentManager#setOffline:this.regions.remove(regionInfo)) 1.3 server C start to open region A.(Not completed) for step3: master ServerShutdownHandler#process() for server B { .. splitlog() ... List regionsInTransition = this.services.getAssignmentManager() .processServerShutdown(this.serverName); ... Skip regions that were in transition unless CLOSING or PENDING_CLOSE ... assign region } {code} In fact, when running ServerShutdownHandler#process()#this.services.getAssignmentManager().processServerShutdown(this.serverName), region A is in RIT (step1.3 not completed), but the return List regionsInTransition doesn't contain it, because region A has removed from AssignmentManager.regions by AssignmentManager#setOffline in step 1.2 Therefore, region A will be assigned twice. Actually, one server killed and started twice will also easily cause region assigned twice. Exclude the above reason, another probability : when execute ServerShutdownHandler#process()#MetaReader.getServerUserRegions ,region is included which is in RIT now. But after completing MetaReader.getServerUserRegions, the region has been opened in other server and is not in RIT now. In our testing environment where balancing,moving and killing are executed periodly, assigning region twice often happens, and it is hateful because it will affect other test cases. was: Before assigning region in ServerShutdownHandler#process, it will check whether region is in RIT, however, this checking doesn't work as the excepted in the following case: 1.move region A from server B to server C 2.kill server B 3.start server B immediately Let's see what happen in the code for the above case {code} for step1: 1.1 server B close the region A, 1.2 master setOffline for region A,(AssignmentManager#setOffline:this.regions.remove(regionInfo)) 1.3 server C start to open region A.(Not completed) for step3: master ServerShutdownHandler#process() for server B { .. splitlog() ... List regionsInTransition = this.services.getAssignmentManager() .processServerShutdown(this.serverName); ... Skip regions that were in transition unless CLOSING or PENDING_CLOSE ... assign region } In fact, when running ServerShutdownHandler#process()#this.services.getAssignmentManager().processServerShutdown(this.serverName), region A is in RIT (step1.3 not completed), but the return List regionsInTransition doesn't contain it, because region A has removed from AssignmentManager.regions by AssignmentManager#setOffline in step 1.2 Therefore, region A will be assigned twice. {code} Actually, one server killed and started twice will also easily cause region assigned twice. Exclude the above reason, another probability : when execute ServerShutdownHandler#process()#MetaReader.getServerUserRegions ,region is included which is in RIT now. But after completing MetaReader.getServerUserRegions, the region has been opened in other server and is not in RIT now. In our testing environment where balancing,moving and killing are executed periodly, assigning region twice often happens, and it is hateful because it will affect other test cases. Affects Version/s: 0.92.0 > Region would be assigned twice easily with continually killing server and > moving region in testing environment > --- > > Key: HBASE-4899 > URL: https://issues.apache.org/jira/browse/HBASE-4899 > Project: HBase > Issue Type: Bug >Affects Versions: 0.92.0 >Reporter: chunhui shen > Attachments: hbase-4899.patch > > > Before assigning region in ServerShutdownHandler#process, it will check > whether region is in RIT, > however, this checking doesn't work as the excepted in the following case: > 1.move region A from server B to server C > 2.kill server B > 3.start server B immediately > Let's see what happen in the code for the above case > {code} > for step1: > 1.1 server B close the region A, > 1.2 master setOffline for region > A,(AssignmentManager#setOffline:this.regions.remove(regionInfo)) > 1.3 server C start to open region A.(Not completed) > for step3: > master ServerShutdownHandler#process() for server B > { > .. > splitlog() > ... > List regionsInTransition = > this.services.
[jira] [Updated] (HBASE-4899) Region would be assigned twice easily with continually killing server and moving region in testing environment
[ https://issues.apache.org/jira/browse/HBASE-4899?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] chunhui shen updated HBASE-4899: Attachment: hbase-4899.patch > Region would be assigned twice easily with continually killing server and > moving region in testing environment > --- > > Key: HBASE-4899 > URL: https://issues.apache.org/jira/browse/HBASE-4899 > Project: HBase > Issue Type: Bug >Reporter: chunhui shen > Attachments: hbase-4899.patch > > > Before assigning region in ServerShutdownHandler#process, it will check > whether region is in RIT, > however, this checking doesn't work as the excepted in the following case: > 1.move region A from server B to server C > 2.kill server B > 3.start server B immediately > Let's see what happen in the code for the above case > {code} > for step1: > 1.1 server B close the region A, > 1.2 master setOffline for region > A,(AssignmentManager#setOffline:this.regions.remove(regionInfo)) > 1.3 server C start to open region A.(Not completed) > for step3: > master ServerShutdownHandler#process() for server B > { > .. > splitlog() > ... > List regionsInTransition = > this.services.getAssignmentManager() > .processServerShutdown(this.serverName); > ... > Skip regions that were in transition unless CLOSING or PENDING_CLOSE > ... > assign region > } > In fact, when running > ServerShutdownHandler#process()#this.services.getAssignmentManager().processServerShutdown(this.serverName), > region A is in RIT (step1.3 not completed), but the return List > regionsInTransition doesn't contain it, because region A has removed from > AssignmentManager.regions by AssignmentManager#setOffline in step 1.2 > Therefore, region A will be assigned twice. > {code} > Actually, one server killed and started twice will also easily cause region > assigned twice. > Exclude the above reason, another probability : > when execute ServerShutdownHandler#process()#MetaReader.getServerUserRegions > ,region is included which is in RIT now. > But after completing MetaReader.getServerUserRegions, the region has been > opened in other server and is not in RIT now. > In our testing environment where balancing,moving and killing are executed > periodly, assigning region twice often happens, and it is hateful because it > will affect other test cases. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira